基于标准粒子群算法的混凝土箱梁水化热过程热工参数反演

Thermal parameters inversion in hydration thermal process of concrete box girder based on SPSO algorithm

针对通过施工现场和实验室试验获取混凝土箱梁水化热仿真分析所需的热工参数缺乏一定的准确性和便捷性,以某混凝土箱梁水化热过程为试验背景,结合文献研究结果确定混凝土箱梁热工参数的取值范围,采用方差分析确定各参数对温度的敏感性,并通过排序筛选敏感性高的参数作为待反演参数,基于标准粒子群算法,对比遗传算法对敏感性高的5个参数进行反演.研究结果表明:混凝土箱梁浇筑过程中,水泥水化热对温度的影响最大,智能算法能有效反演混凝土箱梁热工参数;当迭代次数增大到一定的程度时,标准粒子群算法对应的目标函数小于遗传算法对应的目标函数,遗传算法收敛过程曲线比较平缓,而标准粒子群算法的早期有突变.

To improve the accuracy and convenience of obtaining thermal parameters for hydration thermal simulation analysis of concrete box girder by field and laboratory tests, the hydration heat process of a concrete box girder was used as experimental background. The literature research results were combined to determine the range of thermal parameters of concrete box girder, and the variance analysis was used to determine the sensitivity of each parameter to temperature and sort the parameters with high sensitivity as the parameters to be inverted. Based on standard particle swarm optimization(SPSO) algorithm and compared with genetic algorithm(GA), 5 parameters with high sensitivity were inverted. The results show that the hydration heat of cement has the greatest influence on the temperature during the pouring process of concrete box girder, and the intelligent algorithm can effectively invert the thermal parameters of concrete box girder. When the number of iterations is increased to a certain extent, the objective function corresponding to SPSO algorithm is smaller than the corresponding objective function of GA, and the convergence process curve of GA is relatively flat, while that of SPSO algorithm has mutation in the early stage.